Most error tracing methods use different levels of tracing according to the level of detail needed. Computer code may include trace messages that are generated when the code is executed. The trace messages are usually put into a log file for later analysis and debugging by a user, such as a software tester. In some cases, the user may designate the level of tracing desired. For example, the tracing level may extend from minimal logging (i.e., logging of a small number of trace messages) to verbose logging (i.e., logging of numerous trace messages).
Usually, the default tracing level is minimal. Once a log is generated, a user analyzes the log by first looking for the time frame the error occurred. However, if more logging information is needed than the minimal default logging provides, then another run may be needed with an increased trace level in order to capture the necessary detailed logging to troubleshoot the error. But the error may not necessary reproduce during the re-run. Also, a re-run is an extra effort and may not be easy to perform. For example, the initial run may have been done on a remote location, in a different time zone, etc. Further, the second (verbose) log generated will contain redundant information that is not needed to analyze the actual error and only pollute the log making the troubleshooting process more difficult.